人性化奥林匹克智能体：一种结合求解、评审、维基、验证的多会话奥林匹克问题解决工作流
Humanize Olympiad Agents: A Multi-Session Workflow for Olympiad Problem Solving with Solve, Review, Wiki, Verify
AI总结：
HOA是结合求解、评审、维基、验证的开放多会话框架，通过三类反馈通道提升LLM智能体奥林匹克问题解决能力，在闭源与开源模型下均取得优异成绩，可缩小基础模型能力差距。
摘要：
大型语言模型（LLM）智能体在长周期科学与奥林匹克任务上的表现正快速提升。近期系统已完成纳维-斯托克斯问题的Lean形式化解法，以及Erdős猜想数据库中4个开放问题的自主解答，但两类系统均未同时发布其底层模型与完整推理栈。更广泛而言，单一智能体会话在难题上仍存在脆弱性：早期错误会传播，缺失的领域知识会阻碍进展，表面成功可能掩盖对原始任务的误解或弱化。我们提出HOA（Humanize Olympiads Agent，人性化奥林匹克智能体），这是一个开放的流级框架，通过三个互补反馈通道协调多个智能体会话：第一，独立评审员评估每个候选方案并提供可操作反馈；第二，结构化维基提供模型预训练中可能缺失或不可靠的任务特定知识；第三，验证器对照可信挑战声明检查生成的Lean证明。使用GPT-5.6 Sol时，HOA在2026年IMO、IPhO、IOI中取得满分，在2026年IChO试题中接近满分，在PutnamBench的672道题中全部答对（100%）；在Lean-Eval排行榜上，HOA排名第一，解决了304道题中的251道（82.6%）。我们还使用开源模型评估HOA：使用Kimi-K3时，HOA在IMO、IPhO、IOI中取得满分，在IChO中取得97.1%的金牌级分数。这些结果与闭源GPT-5.6 Sol + /Goal基线在IMO和IPhO上的表现接近，同时在IChO和IOI上超过该基线，证明系统性反馈可帮助弥合基础模型能力的显著差异。
英文摘要：
Large language model (LLM) agents are rapidly improving on long-horizon scientific and olympiad tasks. Recent systems have produced a Lean-formalized resolution of the Navier--Stokes problem and autonomous solutions to four open questions in the Erdős Conjectures database. Yet neither system releases both its underlying model and complete inference stack. More broadly, a single agent session remains brittle on difficult problems: early errors can propagate, missing domain knowledge can stall progress, and apparent success can mask a misinterpretation or weakening of the original task. We introduce HOA (Humanize Olympiads Agent), an open, flow-level framework that coordinates multiple agent sessions through three complementary feedback channels. First, an independent reviewer evaluates each candidate and provides actionable feedback. Second, a structured wiki supplies task-specific knowledge that may be missing or unreliable in model pretraining. Third, a verifier checks generated Lean proofs against trusted challenge statements. With GPT-5.6 Sol, HOA achieves perfect scores on the IMO, IPhO, and IOI 2026, near-full marks on the IChO 2026 problem set, and solves all 672 problems (100%) on PutnamBench. On the Lean-Eval leaderboard, HOA ranks first, solving 251 of 304 problems (82.6%). We further evaluate HOA with open models. Using Kimi-K3, HOA achieves full scores on the IMO, IPhO, and IOI, and a gold-level score of 97.1% on the IChO. These results closely match the closed-source GPT-5.6 Sol + /Goal baseline on the IMO and IPhO, while exceeding it on the IChO and IOI, demonstrating that systematic feedback can help bridge substantial differences in base-model capability.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
